The negative effects of the electronic collar on the welfare of dogs and positive training methods as alternatives
Magazine for Veterinary Science
The government has announced a ban on electronic collars, leading to public debate that will also involve veterinarians. What is the behavioural and physical impact of using electronic collars? Can veterinarians support and substantiate the ban on electronic collars? What can veterinarians recommend as an alternative to using electronic collars for dogs with undesirable behaviour?
An estimated 1.5 million dogs live in the Netherlands, some of which follow obedience courses at one of the 600 dog schools in the Netherlands. Other dogs, mainly working dogs, follow training programmes to become police dogs, tracker dogs or hunting dogs at training groups such as the Royal Dutch Police Dogs Association (KNPV), the Dutch Association for Utility Dogs (NBG) and the Royal Dutch Hunters’ Association (KNJV).
During these training courses or programmes, some instructors or trainers use tools such as electronic collars (also called electric shock collars) or correction chains. The pinch collar, a slip chain with inward-pointing prongs, has been officially banned in the Netherlands since 1 July 2018. Private individuals also sometimes use electric shock collars for undesirable behaviour. A French survey of 1251 dog owners found that 26% of respondents had used an electronic collar at one point or other. Of that 26%, 71.8% had not sought professional advice regarding the effects of such a collar beforehand. In certain countries, including Germany, Denmark, Sweden, Switzerland, Luxembourg, the Czech Republic, Romania and Cyprus, the use of electric shock collars is prohibited or restricted. The Dutch government has similarly announced a ban on the use of this type of collar in the Animals Act.
Types of electronic collars and how they work
An electronic collar is equipped with a battery and a two-pole electrical circuit, which can deliver an electric shock to the neck of the dog. There are three types of electronic collar: 1) the 'training collar' with a remote control for the trainer (also known as the 'dressage collar'), 2) the 'anti-bark collar', which reacts automatically to vibrations of the vocal cords and 3) the electronic fence, where the collar is linked to an invisible fence to keep a dog within a restricted area. Some collars emit a vibration or beep as a warning signal prior to the shock, allowing the animal to avoid the shock by not performing the undesirable behaviour. However, the dog must have previously experienced that the signal is followed by a shock and therefore learned that the signal is a predictor of the shock.
The effect of electronic collars is mainly based on the fact that they cause an aversive stimulus (electric shock) that can also cause physical pain. With excessive use, the electrodes can even lead to lesions on the neck, see image below. In France, where electronic collars are not banned, 7% of the 330 dogs that were fitted with an electric shock collar developed physical injuries. The effect of an electric shock depends on the characteristics of the stimulus, such as its strength, the number of times it is given and its duration.
Studies conducted in humans show that higher frequencies (30, 60 and 90 Hz) cause stronger pain sensations and shock reactions than lower frequencies. Furthermore, studies conducted in humans have shown that successive shocks in the same place increase the sensitivity to pain, while the behavioural startle response may decrease.
Research into the treatment of barking behaviour in dogs shows that several shocks are often necessary before barking behaviour directed at other dogs is completely eliminated. Multiple, strong stimuli are needed to stop the behaviour, particularly where a dog is highly motivated to perform such behaviour. In addition, an electronic collar, such as an anti-bark collar, can sometimes be activated by other (sound) vibrations than a bark or may be activated by another barking dog.
How an individual experiences the electric shock also depends on individual physical characteristics, such as thickness and moisture of the skin and the fat content of the fur. In practice, since many factors are involved, it is very complex to adjust an electric shock to the individual in terms of strength, duration and the number of times the shock is applied to avoid causing any mental trauma. In general, it appears that it is difficult to correctly apply an electronic collar, meaning that the risks of adverse effects, including side-effects, are higher. For example, if the dog associates the electric shock with the owner, this can have a short or long-term negative impact on the relationship with the owner. Furthermore, administering an electric shock can potentially lead to redirected aggression.
Conditioning using the electronic collar
The training of dogs is largely based on the principles of operant conditioning. This means that animals learn to make an association between their behaviour (action) and a certain consequence. That consequence may be a reward such as a treat, which increases the likelihood that the behaviour will occur in the future (positive reinforcement). If the consequence is a punishment such as an electric shock, the probability of the behaviour occurring in the future decreases (positive correction). Alternatively, a stimulus with a positive value or a negative value for the animal can be removed (negative correction or negative reinforcement respectively). For an illustration of these basic principles, please see the figure below.
Training with the help of an electronic collar takes place by means of positive correction or negative reinforcement. In the case of positive correction, the dog receives an electric shock immediately after exhibiting the undesired behaviour. When the electric shock is used as negative reinforcement, the dog is given a shock until it exhibits the desired behaviour. Even if the dog can avoid the shock by way of a predictive signal that precedes the shock, this counts as negative reinforcement. In the literature, the option of using electronic collars as negative reinforcement (avoidance learning) is seen as a training advantage. However, it is not, as avoidance learning is caused by the threat of an unpleasant stimulus. This often means that the dog wears the collar for a long period of time in order to (permanently) suppress the undesired behaviour in different situations. After all, if the threat of electric shock is absent, there is no longer any reason for the dog not to carry out the undesirable behaviour. This leads to the dog having to wear the electronic collar permanently for a long period of time and thus constantly experiencing the threat of a possible electric shock. Both positive correction and negative reinforcement using an electronic collar are based on a strongly aversive stimulus, or the expectation thereof, and belong to the aversive training categories. Several studies show that aversive training is correlated with parameters of reduced well-being.
Welfare effects of electric shock
An animal is in a state of well-being when it is able to actively adapt to its living conditions and thereby achieve a state that it experiences as positive. Over the course of their lives, animals are exposed to a variety of stimuli that induce acute stress, to which they can normally adapt through behavioural and physiological responses.
However, stimuli can be so aversive in terms of intensity and duration that an individual is eventually no longer able to adapt, which results in its welfare being compromised. Research in dogs shows that the use of the electronic collar leads to behavioural and physiological changes associated with stress, (redirected) aggression, fear or even pain. Several studies on Beagles have shown that a shock from an electronic collar when it is not predictable and/or controllable for the dogs leads to a significant increase in the stress hormone cortisol in their saliva.
Working dogs trained with an electronic collar showed significantly more stress-related behaviour and lower postures, as well as vocalisations indicating pain, compared to dogs trained with the slip chain and pinch collar. It is well known, as a result of research into stress physiology, that the degree of controllability and predictability of an unpleasant stimulus is crucial in relation to the amount of stress an animal experiences. Repeated exposure to unpredictable and uncontrollable aversive stimuli, where no behavioural modification leads to a reduction of the stressful stimulus, can lead to deterioration in welfare and even to an animal becoming totally passive. This phenomenon, 'learned helplessness', was demonstrated in dogs by Seligman & Maier (1967).
Dogs that had previously been exposed to an uncontrollable shock no longer showed escape responses, even in situations other than those in which they were given an electric shock. In practice, dogs usually wear an electronic collar for a long time. Certainly at the outset, it is not clear to the dog when and for what reason it is being corrected. This must be learned through trial and error. This makes the electric shock unpredictable, particularly in the beginning. In addition, electronic collars are used to teach or unlearn different types of behaviour. This makes the electric shock even less predictable for dogs, as it is harder for them to link the shock to their own behaviour. Studies show that working dogs may not associate the shock with their own behaviour, but rather associate the shock with the owner, the environment or a command. If the dog does not associate the shock with its behaviour, it will not understand what it must do to avoid the shock.
Inconsistent use and/or incorrect timing further contribute to increased unpredictability and/or uncontrollability. An electric shock is such a strong aversive stimulus that it can quickly cause fear or mental trauma. The administration of one or more electric shocks, or the long-term inability to avoid these shocks, simulates symptoms in rats that serve as animal models for research into human anxiety disorders such as post-traumatic stress disorder (PTSD) and depression. Fear is a functional, behavioural response to a stressor and contributes to an individual's survival. However, constantly responding to a stimulus with very intense fear behaviour, without this leading to a more positive state, is not functional. Anxiety can spread to other stimuli to which the animal did not respond with fear before (generalisation) and, if the anxiety is excessive, dysfunctional and not adaptive or in response to a stimulus, develop into pathological anxiety.
Alternatives: positive training
An electronic collar can effectively suppress undesired behaviour, even in the case of self-rewarding behaviour such as chasing, but it can demonstrably affect a dog's welfare and possibly even negatively affect the relationship with the owner. However, training without aversive methods can lead to better performance, improved welfare, fewer behavioural problems, a better human-animal relationship and improved obedience. A recent publication in the Diergeneeskundig Memorandum provides veterinarians with information on how to prevent unwanted behaviour or redirect it towards desired behaviour without the use of aversive means, by altering the management, modification and/or motivation of the behaviour. In the case of undesirable behaviour, a referral can also be made to a well-trained professional animal behaviourist who works on the basis of the principles of positive training.
Deldalle and Gaunet (2014) demonstrated that domestic dogs that were trained using positive reinforcement (n=24) showed significantly less stress-related behaviour, had more elevated postures and looked at their owners more often than dogs trained using negative reinforcement (n=26). In addition, Schalke et al. (2010) showed that it is possible to train working dogs solely on the basis of positive reinforcement. The study carried out by Cobb et al. (2015) indicates that factors such as selection, socialisation, housing and training determine the welfare and performance of working dogs. The shift towards positive training of working dogs requires a broad approach with regard to: 1) the selection of working dogs (with less excitement), 2) the gradual building up of training from an early age through positive reinforcement (aimed at self-control and letting go) and 3) focus on effective socialisation and mother-child bonding (to stabilise the stress system).
Electronic collars are based on the use of pain stimuli to stop undesired behaviour (positive correction) and/or to teach dogs to exhibit the desired behaviour by making them avoid painful stimuli (negative reinforcement). The electronic collar is an aversive tool, and aversive training is associated with physiological and behavioural parameters of reduced animal welfare. Training dogs without aversive means, through positive reinforcement, is better for the animal's welfare, provides equally good and possibly even better results and leads to a better human-dog relationship. For working dogs specifically, this requires changes in the breeding selection, a change in the approach and structure of the training and focus on effective socialisation and mother-child bonding. There are many indications in the literature that justify a ban on electronic collars. It is critical that veterinarians support the ban on the electronic collar, as they are able to inform dog owners about the detrimental welfare effects of this collar and of the alternatives available in the form of positive training methods.
Published in the Tijdschrift voor Diergeneeskunde (Netherlands Journal of Veterinary Science) in September 2020.
Authored by: C. M. Kapteijn, J.A.M. van der Borg, C.M. Vinke, S.S. Arndt Animal Behaviour Research Group, Animals in Science and Society unit, department of Population Health Science, Faculty of Veterinary Medicine, Utrecht University.